Vienna, Austria

ESTRO 2023

Session Item

Monday
May 15
16:30 - 17:30
Business Suite 3-4
Treatment planning: Photons
Sara Pilskog, Norway
Poster Discussion
Physics
Mid-ventilation vs ITV for SABR on the MR-Linac for locally advanced pancreatic cancer
Jessica Gough, United Kingdom
PD-0983

Abstract

Mid-ventilation vs ITV for SABR on the MR-Linac for locally advanced pancreatic cancer
Authors:

Jessica Gough1,2, Robert Adam Mitchell3, Alex Dunlop3, Dualta McQuaid1, Simeon Nill3, Sophie Alexander1,2, Helen McNair1,2, Uwe Oeflke3, Brian Ng-Cheng-Hin1, Arabella Hunt1, Katharine Aitken1,2

1Royal Marsden Hospital, Radiotherapy, Sutton, United Kingdom; 2Institute of Cancer Research, Radiotherapy and Imaging, London, United Kingdom; 3The Royal Marsden Hospital and the Institute of Cancer Research, Joint Department of Physics, London, United Kingdom

Show Affiliations
Purpose or Objective

For Stereotactic Ablative Body Radiotherapy (SABR), it is desirable to limit target volumes to deliver ablative doses whilst sparing surrounding organs at risk (OARs). Internal target volumes (ITVs) may be used to account for respiratory motion if advanced motion management methods are not available. Delivering escalated doses may improve survival for locally advanced pancreatic cancer (LAPC), but safe SABR delivery requires limiting dose to adjacent GI OARs. Mid-ventilation (MidV) PTV margins have been shown to be dosimetrically advantageous for delivering conventionally fractionated RT (CFRT)to LAPC(1). This planning study compared ITV with MidV PTV margins for MR-Linac (MRL) SABR delivery for LAPC.

Material and Methods

10 LAPC patients previously treated with CFRT were identified. For each patient, 2 SABR plans were generated for 35Gy in 5 fractions: an ITV and MidV plan. For MidV plans, the MidV phase (used for planning) was determined by contouring the GTV on all 10 4DCT phases and identifying the phase where the GTV centre of mass was closest to the average position. Individualised PTVMidV margins were calculated for each plane according to a modified Van Herk recipe(1,2)where motion was incorporated as a random error component. For ITV plans, the GTV was encompassed through its entire trajectory to form the ITV and a 3.1mm isometric PTVITV margin applied. PTVITV plans were produced on contrast enhanced CT. All error values within the PTVITV margin formula were the same as those used for PTVMidV margin calculations. IMRT plans were generated using Monaco(v5.40, Elekta, Stockholm). PTV coverage was optimised whilst respecting mandatory OAR constraints. OAR constraints were as per UK LAPC SABR guidelines(3). PTV volumes and PTV/OAR dosimetry were compared and assessed using Wilcoxon matched pairs test (significant p=≤0.05).

Results

PTVMidV volumes were significantly smaller than PTVITV volumes (median 84.86cm³ vs 108.07cm³, p=0.002). Dosimetric parameters are shown in Table 1. Larger PTVITVs resulted in more overlap with normal tissues, requiring greater compromise of PTV coverage to meet OAR constraints compared to MidV plans. For 1 patient, a clinically acceptable SABR plan was not achievable with the PTVITV approach, as adequate PTV coverage (V35Gy ≥ 60%, as per UK guidelines(3)) could not be achieved without exceeding mandatory OAR constraints. For this case, V35Gy:PTVITV = 54.9%, PTVMidV= 81%. The MidV approach always resulted in acceptable plans.



Conclusion

Use of PTVMidV margins resulted in significantly smaller PTV volumes with less OAR overlap and a modest improvement in PTV coverage compared with PTVITV margins. A PTVMidV approach may help to produce clinically acceptable SABR plans (where not possible with PTVITV) and could facilitate dose escalation. We recommend PTVMidV margins for MRL SABR for LAPC and plan to use this approach in combination with abdominal compression. This will be reassessed when advanced motion management is available on the Elekta Unity MRL.